Run below script at the host with root account:
for i in /*; do echo $i; find $i |wc -l; done
Once identified certain directory included big amount of files, then we can dig down to see which base directory caused the problem:
for i in /xxx/xxx/*; do echo $i; find $i |wc -l; done
If the related directory's file can be removed, we can use below command to delete them
cd /xxx/xxx/; find . -type f | xargs rm -rf
At last, check inode usage again.
df -i /xxxx/
Note: Removing files from your segment data directory can leave the segment in an inoperable state and may cause permanent data loss. If you encounter this error and cannot free enough inodes outside of the segment data directories, please contact support.